Authentication of Lambs Feed under Dry and Hot Environment Based on Color Parameters of Fats from Fat-tailed and Thin-tailed Breeds’ Measured Using Visible Reflectance Spectroscopy

Considering the additional market value of pasture meat many authentication methods were developed to discriminate it from meat produced in conventional systems. The visible reflectance spectroscopy technique has proven to be effective in temperate European conditions and breeds. This study aimed to apply reflectance spectroscopy technique to authenticate lamb meat from grassland-based systems under North African conditions. This work tested the reliability of this method to discriminate between pasture-fed (P) and stall-fed (S) lambs under North African conditions and investigated the effect of feeding system (FS) (P vs. S) and breed (Barbarine, BB; Queue Fine de l’Ouest, QFO and Noire de Thibar, NT) on weight and color of perirenal, subcutaneous and caudal fat. There were a total of 18 P and 18 S lambs used, with 6 P and 6 S lambs used for each breed. The colour and reflectance spectrum of various fat tissues were assessed. The FS affected weights of all fat tissues and allcolor parameters of perirenal and subcutaneous fat (P 0.01); almost it affected redness and yellowness of caudal fat (P 0.05; P 0.01). S lambs had higher lightness and lower redness and yellowness in all adipose tissues than P lambs. Weight, lightness, and redness of perirenal fat, as well as weight and redness of subcutaneous fat, were all affected by the breed, with a significant interaction with FS for subcutaneous fat data. The reflectance spectrum of perirenal, subcutaneous, and caudal fat at wavelengths between 450 and 510 nm (method 1, M1) or at wavelengths between 400 and 700 nm using partial least squares discriminative analysis (PLS-DA) as a classification method (method 2, M2) were used to differentiate P lambs from S lambs. M2 yielded to a higher proportion of correctly classified lambs compared with M1 (P = 0.001). The proportion of correctly classified lambs using M2 was 76.4, 75.0 and 80.0 % for perirenal, subcutaneous and caudal fat for P lambs and 83.3, 76.4 and 100.0 % for perirenal, subcutaneous and caudal fat for S lambs. Despite lower reliability when compared to European studies, this study confirmed the efficacy of the visible reflectance spectroscopy technique used on perirenal fat in feeding system authentication under North African conditions and identified caudal fat as a new support for better fat-tailed breed classification.